Reading the EMS Literature
A major goal of PHTLS has been to ensure that the practice recommendations presented in this text accurately represent the best medical evidence available at the time of publication in support of these recommendations. To that end, PHTLS began this process with the 6 th edition and it is expanded in the 7th edition . We continue to add, as References and Suggested Readings, those manuscripts , sources, and resources that are fundamental components of each chapter and recommendation. (See Suggested Readings at the end of this chapter for further information on evaluating the EMS literature.) Every practitioner and health care provider should obtain, read, and critically evaluate the publications and sources that make up the basis for all the components of daily practice. To accomplish this, an understanding of exactly what constitutes the medical literature and of how to interpret the various sources of information is essential. In many cases the first place that is accessed for information about a particular topic is a medical textbook. As our level of interest and sophistication grows, a search is undertaken to find the specific references that are referred to in those book chapters or to find what, if any, primary research studies have been performed and published. Then, after reviewing and analyzing the various sources, a decision can be made about the quality and strength of the evidence that will guide our decision making and patient care interventions.
Types of Evidence
There are a number of different systems for rating the quality and strength of medical evidence. Regardless of the exact rating system used, several common assessments can be found among them. The highest-quality source that leads to the strongest recommendation about the treatment under study is the randomized, double-blind controlled study. Studies of this type are usually referred to as Class I evidence. This type of study is considered to be the best type of study because: all patients entered into the study are randomized (meaning each patient has an equal chance of assignment to whatever type of treatment is being studied); the researchers as well as the patients do not know which type of treatment the subject is receiving (double blinding); and investigators are controlling as many other aspects of the study as possible. These factors minimize the chances of any bias entering into the study or affecting the results or interpretation of the results. Class II evidence generally includes the other types of studies that can be found in the medical literature, including nonrandomized non-blinded studies, retrospective case-control series, and cohort studies. Finally, Class III evidence consists of case studies , case reports, consensus documents, textbook material, and medical opinion. Class III evidence is the weakest source of evidence, although often the easiest to obtain. Unfortunately, if the literature related to prehospital care is critically reviewed, the majority qualifies as Class III evidence. There has been remarkably little research that would qualify as Class I. Much of the practice of medicine that has been applied to the prehospital setting has been adopted from and adapted to the out-of-hospital environment from the in-hospital delivery of emergency care. The result is that most of the prehospital care provided today is based on Class III evidence. However, more and more Class I and II studies related to prehospital care are being conducted. Unfortunately, Class I studies are limited in the United States by rigid informed-consent regulations. Specifically, with few exceptions, the medical practice of prehospital care is based on “expert” opinion usually found in textbook chapters, the credentials and qualifications of the individual offering that opinion, and the forcefulness and “volume” of the delivery of that opinion. type of treatment the subject is receiving (double blinding); and investigators are controlling as many other aspects of the study as possible. These factors minimize the chances of any bias entering into the study or affecting the results or interpretation of the results. Class II evidence generally includes the other types of studies that can be found in the medical literature, including nonrandomized non-blinded studies, retrospective case-control series, and cohort studies. Finally, Class III evidence consists of case studies , case reports, consensus documents, textbook material, and medical opinion. Class III evidence is the weakest source of evidence, although often the easiest to obtain. Unfortunately, if the literature related to prehospital care is critically reviewed, the majority qualifies as Class III evidence. There has been remarkably little research that would qualify as Class I. Much of the practice of medicine that has been applied to the prehospital setting has been adopted from and adapted to the out-of-hospital environment from the in-hospital delivery of emergency care. The result is that most of the prehospital care provided today is based on Class III evidence. However, more and more Class I and II studies related to prehospital care are being conducted. Unfortunately, Class I studies are limited in the United States by rigid informed-consent regulations. Specifically, with few exceptions, the medical practice of prehospital care is based on “expert” opinion usually found in textbook chapters, the credentials and qualifications of the individual offering that opinion, and the forcefulness and “volume” of the delivery of that opinion. Recently, consensus seems to be building to utilize a system for grading the quality of evidence and the strength of a resulting clinical practice recommendation known as The Grades of Recommendation, Assessment, Development and Evaluation (GRADE). This system appraises the quality of evidence available and the benefits versus the risks of the intervention, and then develops a judgment about the strength of a particular treatment recommendation or intervention. Evidence is rated into one of four categories: high, moderate, low, very low quality. Based upon the quality of the evidence and the benefit/ risk evaluation, the resulting recommendation is categorized as strong or weak.
Steps in Evaluation
Every medical practitioner should read the medical literature and critically evaluate every study published that might alter treatment decisions in order to distinguish useful information and therapy from that which is useless or potentially even harmful. How, then, does one go about reading and The first step in this process is to develop a list of journals that will form the foundation of a regular literature review. This list should comprise not only those journals with the desired specialty in their name, but also those publications that address related specialties or topics and have a high likelihood of also publishing applicable studies ( Figure 1-7 ). An alternative to reviewing multiple journals is to perform a computerized literature search if there is a particular topic of interest. The use of computerized search engines such as Medline or Ovid allows the computer to search a massive database of multiple medical journals and automatically develop a list of suggested studies and publications ( Figure 1-8 ). The next step is to review the title of every article in the table of contents of each of the selected journals to narrow down the article choices to those that clearly relate to the topic of interest. It would be impossible to read each of the selected journals from cover to cover, nor is it necessary. By reviewing the table of contents, articles that are of no interest can be immediately dismissed. Once the selection has been narrowed, there are still a number of preliminary actions to take before reading the text of the article. Look at the authors listed on the articles to see if any are already known for their work in this area. Next, read the summary or abstract of the article to see if this overview of the article fulfils the expectations generated when the title was first reviewed. Then, review the site where the study was conducted to assess the similarities, differences, and applicability to the setting in which the results of the study may be subsequently applied. It must be stressed that reading the abstract alone is not enough. It serves only as the “teaser” to determine whether the full article should be reviewed. Medical practice should never be changed based on the abstract. Once these initial points have been evaluated , the full text of the article is now read and critically assessed. In doing so, several specific issues are important to determine. The first issue is to evaluate the study and the randomization of the patients entered into the study. Every patient included should have had the same probability of receiving one or the other treatments or interventions being compared in the study, and that probability should be known beforehand. The method for assigning patients to their treatment should be described and should be similar to the flip of a coin. Next, the patient population entered into the study is assessed to determine the similarities or differences with the target population for which the study conclusions will be implemented. To do so, adequate information must be provided in the text describing the clinical and sociodemographic makeup of the study population. Ideally, studies that will be used to alter the care provided in the prehospital setting should have been performed in the prehospital setting. As Dan Spaite, MD, stated “Strong evidence for efficacy of an intervention does not mean that it will be effective when applied in the field.” The next issue for consideration is the outcome measure selected by the authors. All outcomes that are clinically relevant should be considered and reported in the study. For example, cardiac arrest studies may describe such endpoints or outcomes as cardiac rhythm conversion, return of spontaneous circulation, survival to hospital admission, or survival to discharge from the hospital. Analysis of the results section also requires critical review. Just as it is important to evaluate the study population and entry criteria, it is equally important to see if all patients entered into the study at the beginning are accounted for at the end of the study. Specifically, the authors should describe any criteria used to exclude patients from the study analysis. Simple addition by the reader of the various treatment groups or subgroups will quickly confirm if all patients were accounted for. The authors should also describe occurrences that might introduce bias into the results. For example, the authors should report mishaps such as control patients accidentally receiving the treatment or study patients receiving other diagnostics or interventions. In addition, the clinical, as well as the statistical, significance of the results are important to consider. Although the statistical analyses may be difficult to understand, a basic comprehension of statistical test selection and utilization will validate the statistical tests performed. Equal to and perhaps more important than the statistical significance of a result is the clinical significance of the reported result. For example, in evaluating the effect of a new antihypertensive medication, the statistical analysis may show that the new drug causes a statistically significant decrease in blood pressure of 4 mm Hg. Clinically, however, the reported decrease is insignificant. Thus, the health care provider must assess not only the statistical significance of the result but also the clinical significance. If all the prior issues have been answered satisfactorily, the last issue relates to the implementation of the study results and conclusion in the reader’s health care system. To determine the practicality of applying the therapy, the authors needed to describe the treatment in sufficient detail, the intervention or therapy must be available for use, and it must be clinically sensible in the planned setting. Some differences exist in the evaluation of consensus statements, overviews, and textbook chapters. Ideally, the statement or overview should address a specific, focused question. The authors should describe the criteria used to select the articles included as references, and the reader will determine the appropriateness of these criteria. This in turn will help determine the likelihood that important studies were included and not missed. In addition, the reference list should be reviewed for known studies that should have been included. A high-quality overview or consensus statement will include a discussion of the process by which the validity of the included studies was appraised. The validity assessment should be reproducible regardless of who actually performed the appraisal. Also, multiple studies with similar results help support the conclusions and ultimate decision about whether or not to change current practice. Similar to the review of individual studies , the review and assessment of consensus statements, overviews, and textbook chapters include a determination of whether all clinically relevant outcomes were considered and discussed and whether the results can be applied to the reader’s patient-care setting. This also includes an analysis of the benefits versus the potential risks and harm. The final step in evaluation is to determine when a publication should cause a change in daily medical practice. Ideally, any change in medical practice will result from a study of the highest quality, specifically a randomized, controlled, double -blind study. The conclusion of that study will be based on results that have been critically evaluated, have both statistical and clinical significance, and have been reviewed and judged to be valid. The study should be the best information currently available on the issue. In addition, the change in practice must be feasible for the system planning to make the change, and the benefit of making the change must outweigh the risks. All the medical practice in the out-of -hospital setting should be based on high-quality, Class I evidence that supports practice. As noted, however, most of the EMS literature qualifies as Class III evidence.
No comments:
Post a Comment